Projects are to be carried out in groups with 1-3 students in each group. Below is a list of possible projects to work on, but you can also choose a different project if you have a specific topic in mind.
Urban design For city planning it is important to know how the structure of the city affects things such as population density, and housing prices. In this project, the aim is to estimate how spatial models using various explanatory variables can predict population denstiy or housing prices. Data for the project has been collected by the Spatial Morphology Group at Chalmers.
Segmentation of micro-CT images For the production of medical tablets, it is important to know how the manufacturing process affects the composition. To do this, one first needs to be able to identify the different components in the tablet based on micro-CT images. The goal of this project is to design a method for image segmentation of such images. The data for the project has been collected by AstraZeneca.
Temperature modeling Download a dataset of temperature measurements and design a number of geostatistical models (with different explanatory variables for example) to interpolate the data. Test the model using cross-validation. A possible dataset is http://www.image.ucar.edu/GSP/Data/US.monthly.met/.
Road detection Take images from Google Streetview and design a method that find the road in the images.
Identification of traffic signs Use Google Streetwiew to extract images with and without traffic signs. Design a method to detect the traffic signs with speed limits. If the method finds one, identify the speed limit.
Identification of traffic signs with names of places. An alternative to the previous projecet. Use Google Streetwiew to extract images with and without traffic signs with names of cities. Design a method to detect the traffic signs. If the method finds one, identify the name of the city.
Face detection. Design a method for finding faces in images.
Face recognition Try to recognize faces in a suitable face database available from internet (or in a database that you construct by use of a digital camera).
Optical scanning of documents Find text in images and store as ASCII-code.
Segmentation of MR images Design a method for automatic segmentation of Magnetic Resonance images.
Tracking the football in a series of images of soccer football Try to follow the football in a series of images. Start perhaps by manually identifying the football in the first image. In the sequel use for instance the last image (or the last two or even the last three images) to find the the football in the following image.
Real time gestures The object is to recognize and track hand gestures acquired by a web camera in real time as an alternative input device to a computer.
Identification of houses with swimming pools Use images from Google Earth and statistical image analysis methods to identify houses with swimming pools.
Classification of facial expression. Recognize faces and classify them into suitable categories (happy, sad, neutral, etc.)
Identification of handwritten signatures Let a number n of test persons write m signatures each. Think of a suitable design for acquiring these nm signatures. Use pattern recognition methods to identify the persons from the signatures.
Identification of tram numbers Take photos with and without trams. Design a method to detect trams and classify which number it has.
Find ships i satellite images Use a database of satellite images with and without ships and design a method for finding the ships in the images.
Analysis of mammography images Mammography images of different types (cancer, benign and normal) are available from http://www.eng.usf.edu/cvprg/Mammography/Database.html. Download a suitable number of images for each of the three classes from the database. Divide the data set into a test set and an evaluation set. Try to develop a method for distinguishing between the three classes (or use perhaps only two classes) by use of the test set. Then evaluate the method on the test set.
Gender identification from face images Design a method to classify images of faces according to gender.
Last modified: April 2, 2019